New marine heterotardigrade lineages (Echiniscoididae) from the tropics

Abstract Echiniscoididae are the only family of the clade Echiniscoidea residing in marine habitats. Their characteristic feature is the multiplication of claws on legs, commonly regarded as an adaptation to unstable tidal environment. Although DNA barcoding data indicated the presence of numerous unnamed candidate species within Echiniscoides about a decade ago, only recently have new species been formally described. In this paper, we present new genetic, morphological and morphometric data and establish five Echiniscoides species found in samples acquired from various tropical regions: Echiniscoides basalticus sp. nov. (Mauritius, Indian Ocean), Echiniscoides bufocephalus sp. nov. (Qatar, Indian Ocean), Echiniscoides lichenophilus sp. nov. (Dominican Republic, Atlantic Ocean), Echiniscoides musa sp. nov. (Brazil, Atlantic Ocean), and Echiniscoides trichosus sp. nov. (Rapa Nui, Pacific Ocean). We discuss the taxonomic importance of cuticular sculpturing in Echiniscoididae. Finally, we elevate all subspecies of Echiniscoides sigismundi to species level. The key to all echiniscoidids is provided. http://www.zoobank.org/urn:lsid:zoobank.org:pub:63154710-C3BA-402A-A11B-C211DAE59CAB


Introduction
Tardigrades inhabit a variety of environments, ranging from mountain peaks (Kristensen 1987) to bathyal zone in oceans (Bussau 1992). Marine tardigrades are usually adapted to specific lifestyles, such as dwelling on the surface of bottom sediments (Jørgensen & Kristensen 2001), parasitising holothurians as epibionts , or clinging to thecae of barnacles in the tidal zone . The last type of lifestyle, entailing commensal and/or parasitic relationships with adult barnacles, characterises the family Echiniscoididae . Its type species, Echiniscoides sigismundi (M. Schultze, 1865), represents one of the first described tardigrade species (Ramazzotti & Maucci 1982). Plate (1889) divided Echiniscus C.A.S. Schultze 1840 into two subgenera: Echiniscus and Echiniscoides, using the number of claws per leg as a discriminative criterion. Echiniscus was defined as grouping species with two (in the larval stage) or four claws (in later life stages), whereas species with seven to nine claws per leg were accommodated in Echiniscoides, at the same time underlying the crucial autapomorphy of echiniscoidid tardigrades: the unprecedented multiplication of claws.
No important contributions to the classification of Echiniscoides were published since the revision of Plate (1889) until Renaud-Mornant (1976) described E. sigismundi polynesiensis from the waters of Mo'orea in the French Polynesia. This paper, together with the fundamental revision of the Echiniscoides variability presented in Kristensen and Hallas (1980), somehow consolidated a trend of establishing many new echiniscoidid taxa as subspecies of E. sigismundi, which was in accordance with the prevailing view of that period, assuming cosmopolitan distributions of most meiofaunal taxa (Cerca et al. 2018). Kristensen and Hallas (1980) introduced the claw formula and type of cuticular sculpturing as decisive taxonomic traits. Furthermore, they indicated some similarities between Echiniscoididae and Coronarctidae, such as the shape of secondary clavae (Renaud-Mornant 1987). Later on, Hallas and Kristensen (1982) proposed the length of cirrus E as a useful criterion in the internal classification of Echiniscoides. The echiniscoidid resemblance to Coronarctidae was raised again by D'Addabbo Gallo et al. (1992), who stressed that the annulation of cirrus E has a plesiomorphic nature and, consequently, should not be used is phylogenetic inference.
In a series of important papers, Faurby et al. (2011Faurby et al. ( , 2012 demonstrated that populations of Echiniscoides collected worldwide are strongly geographically differentiated and the main process responsible for diversification of multiple evolutionary lineages within Echiniscoididae was ecological speciation. Faurby and Barber (2015) corroborated the utility of intertidal Echiniscoides populations in phylogeography and augmented the evidence for numerous undescribed species within the genus (Clausen et al. 2014). Perry and Miller (2015) described the first and the only species to date with a flexible posterior portion of the buccal tube, E. wyethi. Finally, the most recent revisions (Møbjerg et al. , 2020 divided Echiniscoididae into Echiniscoidinae and Isoechiniscoidinae on the basis of strong ecomorphological disparities: the claw number and isonychy (7-13 anisonych vs usually 6 isonych), pillars in the epicuticle (absent vs present), the length of sensory appendages (comparatively short vs long), and inhabited niche (barnacles, algae and lichens in tidal zone vs interstitial species, respectively); the anal system was additionally proposed as a next taxonomic trait worth detailed documentation in the case of Echiniscoididae, and the genus Neoechiniscoides established for species with well-developed lateral anal lobes. Last but not least, Møbjerg et al. (2020) instantiated the family Anisonychidae, pointing out to a long-questioned affinity between Anisonyches and Echiniscoides (Grimaldi de Zio et al. 1987;Fujimoto et al. 2017).
In this paper, we aim at unravelling diversity of Echiniscoides populations sampled in tropical waters. We provide integrative descriptions of five new species, discuss the variability and utility of cuticular sculpturing in echiniscoidid and echiniscoidean systematics, and update the taxonomic key to the genus (Hallas & Kristensen 1982) and its relatives. We express hopes that the studies, which triggered molecular research on Echiniscoides (Faurby et al. 2011(Faurby et al. , 2012 and uncovered unparalleled genetic diversity within the genus, will be followed by integrative data, including both an influx of further DNA barcodes and detailed morphological observations.

Sampling and comparative material
Summary of Echiniscoides specimens processed in different analyses is presented in Table I. Samples of barnacles and lichens were first completely dried, and then re-soaked in distilled water. Isolated animals were either mounted on slides or preserved in 95% ethanol for sequencing and scanning microscopy. Furthermore, we examined representatives of E. sigismundi groenlandicus Kristensen & Hallas, 1980 (7♀♀ and 1♂, collected from barnacles in Qeqertarsuaq, Greenland), E. hoepneri Kristensen & Hallas, 1980 (9♀♀ and 2♂♂, collected from barnacles in Iceland;Faurby et al. 2011), and E. sigismundi hispaniensis Kristensen & Hallas, 1980 (5♀♀ and 2♂♂, collected from barnacles at the entrance to the lagoon in Praia da Barra, Central Portugal at 3 rd April 2009; temperature: 16.5°C, salinity = 35.5‰).

Microscopy and imaging
When present, alive specimens were observed and photographed (using objectives up to ×400 magnification). A fraction of specimens was later mounted in glycerol and Hoyer's medium (also in the polyvinyllactophenol in the case of E. trichosus sp. nov.) on permanent slides. Mounted exemplars were photographed using a DP20 camera on an Olympus BX51 compound microscope with differential interference contrast (DIC) optics, and Olympus BX53 light microscope with phase contrast (PCM), associated with an Olympus DP74 digital camera. Preparations for scanning electron microscopy (SEM) were performed according to Hygum et al. (2016). Individuals were photographed and subsequently stained with osmium tetroxide (OsO 4 ). The specimens were carefully rinsed in distilled water and then dehydrated in a graded series of ethanol and acetone prior to critical point drying. Afterwards, the dry specimens were mounted on aluminium stubs and coated with platinum/palladium alloy or gold and analysed with a JEOL JSM-6335 F Field Emission and Versa 3D DualBeam SEM. All light microscopy figures were 720 P. Gąsiorek and R. M. Kristensen assembled in Corel Photo-Paint X8. For deep structures that could not be fully focused on a single PCM photograph, a series of images were taken every ca. 0.1 μm of vertical focusing and then assembled manually in Corel Photo-Paint into a single deep-focus image. General phenotype drawings were prepared based on in vivo photographs of selected individuals and assembled photographs of holotypes.

Morphometry and terminology
Structures were measured only when oriented properly and not broken or deformed, according to the schemes from Kristensen and Hallas (1980). Echiniscoidid claw formula presents claw numbers on subsequent legs . Body length was measured from the anterior to the posterior end of the body, excluding the hind legs. General heterotardigrade terminology follows Fontoura et al. (2017).

Genotyping and phylogenetics
A Chelex® 100 resin (Bio-Rad) extraction method was used in DNA extraction (Casquet et al. 2012;Stec et al. 2020). Four DNA fragments were sequenced: the small ribosome subunit 18S rRNA, the large ribosome subunit 28S rRNA, the internal transcribed spacer ITS-1 (ITS-2 did not amplify in all species), and the cytochrome oxidase subunit I COI. Various 28S rRNA primers were used in literature to amplify Echiniscoides gene fragments (Møbjerg et al. , 2020; thus, we used all of them to obtain different regions of this marker. All fragments were amplified and sequenced according to the protocols described in Stec et al. (2020); primers and original references for specific PCR programmes are listed in Supplementary Material 1. GenBank accession numbers for all newly sequenced species are provided in Table II. All echiniscoidid COI sequences available in GenBank were aligned with Echiniscus testudo (Doyère, 1840) as an outgroup, using the ClustalW Multiple Alignment tool (Thompson et al. 1994) implemented in BioEdit (Hall 1997). We deliberately did not concatenate COI and 28S rRNA sequences since the latter represent various DNA regions and restricting the dataset to echiniscoidid taxa with both available markers would greatly diminish it. There are no available ITS sequences for echiniscoidids, and 18S sequences are accessible for <5% of taxa with COI. That is why we performed a monolocus reconstruction. The aligned fragment  (Lanfear et al. 2017) with applied Bayesian Information Criterion (BIC) and greedy algorithm (Lanfear et al. 2012), the best substitution model and partitioning scheme was chosen for posterior phylogenetic analysis. As the best-fit partitioning scheme, PartitionFinder suggested one partition characterised by TIM+I+G model. Bayesian inference (BI) marginal posterior probabilities were first calculated using MrBayes v.3.2 (Ronquist & Huelsenbeck 2003). Random starting trees were used, and the analysis was run for ten million generations, sampling the Markov chain every 1000 generations. An average standard deviation of split frequencies of <0.01 was used as a guide to ensure that the two independent analyses had converged. Tracer v1.3 (Rambaut et al. 2014) was then used to ensure Markov chains had reached stationarity and to determine the correct "burn-in" for the analysis, i.e. the first 10% of generations. The Effective Sample Size values were greater than 200 and the consensus tree (visualised in FigTree v.1.4.3 available from https://tree.bio.ed. ac.uk/software/figtree) was obtained after summarising the resulting topologies and discarding the "burn-in".

Etymology
The name underlines the substrate (magmatic rock), which contained a barnacle population inhabited by the new species. An adjective in nominative singular.  New marine heterotardigrade lineages from the tropics  Buccal tube 40.9 μm long in the holotype (buccal tube length/body length ratio = 0.22), rigid and wide (2.6 μm). Large stylet furcae dissolved in Hoyer's medium; stylet supports absent. Pharynx with poorly sclerotised placoids ( Figure 4).
Dorsum sculptured with irregular polygonal tubercles (∼1.0-2.0 μm in diameter) that are poorly elevated above the remaining cuticle and faint in PCM. The sculpturing covers only distal part of the head, but does not reach its proximal part, legs (it vanishes at the limb bases) and venter ( Figure 3). All legs with sensory organs: poorly marked and weakly protruding papillae on legs I and II; thin and spine-like on legs III; large papillae on legs IV, differing from primary clavae by the lack of distal pointing tip.

Etymology
From Latin Bufo = toad + Ancient Greek Κέφαλος (kephalos) = head, meaning "toad-headed", which refers to the unique shape of the cephalic body portion in the new species. An adjective in nominative singular.  Figure 7D).

Description
Dorsum sculptured with strong irregular wrinkles that are perpendicular to the main body axis, bestdelineated in the caudal body portion at the level of legs III (Figures 5-7A). Legs and venter smooth. All legs with sensory organs: poorly marked and weakly protruding papillae on legs I; thin and spine-like on legs II (Figures 6, 8D), more robust and spine-like on legs III (Figures 6, 8E); large papillae on legs IV (Figures 6-7A), identical in shape to primary clavae.
Adult males. Found only among specimens processed for SEM. Qualitatively alike females, beside of the circular gonopore with a semicircular slit ( Figure 9C,D).

Differential diagnosis
There are few Echiniscoides species with spine-like sensory organs on legs II, and only one of them, E. ritavargasae, exhibits sculptured dorsal cuticle, but it can be easily told apart from E. bufocephalus sp. nov. based on the type of sculpturing (circular tubercles instead of strong irregular wrinkling), adult claw formula 9,9,9,9, short, tufted cirri externi (2.

Etymology
From Ancient Greek λειχήν (leíkhō = to lick, liverwort) + φῐλος (phílos = love) = "loving lichens". The name alludes to the substrate in which the new species was found. An adjective in nominative singular.
Adult males. Body slenderer and shorter than in females. Claw formula 8/10,8/10,8/9,7/8; in general, males characterised by the body size comparable to that of females have 1-2 claws less per limb. Gonopore oval and weakly protruding. No other differences with regard to females.

Differential diagnosis
There is only one species of Echiniscoides exhibiting papilliform sensory organs on legs III -E. travei Bellido & Bertrand, 1981, but it has smooth cuticle, its representatives attain much larger body size (up to 420 μm), and sensory organs on legs I are not identifiable under PCM.

Description
Adult females. Body cylindrical and elongated ( Figures 14A, 15). Large black eyes present ( Figures 14A, 15). Cirri interni and externi conical, of similar length and with tufted tips. Median cephalic cirrus absent. Lateral cirri A and E short and of similar length, with smooth base and pointed flagellum ( Figures 14A, 15). Papillary primary clavae and lenticular secondary clavae present ( Figures 14A,B,  15). Subterminal mouth without mouth plates. The holotype is in the simplex stage.  Dorsum strongly sculptured with irregular polygonal tubercles (∼1.0-2.2 μm in diameter) that are well-elevated above the remaining cuticle and easily identifiable under PCM ( Figure 14A,B). The sculpturing covers head, but dorsolateralmost body portions, legs and ventral side are smooth ( Figures 14A,B, 15). All legs with sensory organs: poorly marked and weakly protruding papillae on legs I and II; thin and spine-like on  New marine heterotardigrade lineages from the tropics legs III; large papillae on legs IV of the shape analogous to primary clavae ( Figures 14A,B, 15). Anisonych claws ( Figure 14D); claw formula 7,6,7,6. Gonopore hexapartite, rosette-shaped. Anus trilobed, with two lateral lobes and a short posterior lobe, lacking additional lobes.
Juveniles. Not found.

Differential diagnosis
With the adult claw formula 7,6/7,6/7,6, E. musa sp. nov. is similar to only two other species, but it is easily distinguishable from: • E. costaricensis, because the latter exhibits higher conspicuousness of the dorsal sculpture (more elevated granules), which covers legs, and the annulated bases of cirri A and E. • E. hispaniensis, because the latter has different dorsal tubercles when seen in light microscope (arranged like roofing tiles) and sensory organs on legs I-II not identifiable in PCM.
Dorsum seems finely punctuated under PCM (Figure 16), but it consists of unique, minute cuticle protrusions called here microtrichia ( Figures 18B, 19). They are well-identifiable under SEM (Figures 19, 20A-C), and their density is highest in the caudal body portion, gradually becoming less numerous towards head ( Figure 17). The sculpturing does not cover head ( Figure 18C-E), legs and ventral body side. All legs with sensory organs: poorly marked and weakly protruding papillae on legs I and II (Figures 16-17, 18E); thin and spine-like on legs III (Figures 16-17, 18A, 20D); large papillae on legs IV of the shape analogous to primary clavae (Figures 16-17).
Adult males. Body of similar size and shape to that of females. Gonopore oval and weakly protruding. Sexual dimorphism weakly marked.
Juveniles. Not found.

Differential diagnosis
With the adult claw formula 11,11/13,11/13,10, E. trichosus sp. nov. is similar to only two other species, but it is easily distinguishable from: • E. mediterranicus  because the latter exhibits smooth or wrinkled cuticle, and sensory organs on legs I-II are not identifiable under PCM. • E. travei, because the latter has papillary sensory organs on legs III, smooth cuticle, much larger   Updated shortened description Adult females. Body plump ( Figures 21A,B, 21D). Cirri interni and externi conical, of similar length and with weakly tufted tips ( Figure 22A-C). Median cephalic cirrus absent. Lateral cirri A and E short and of similar length, with annulated base and pointed flagellum ( Figure 22D-F). Papillary primary clavae and lenticular secondary clavae present ( Figure 22C,D). Subterminal mouth without mouth plates. Dorsum with flat epicuticular granules ( Figure  22F) in SEM that become poorly elevated in the proximal body part ( Figure 21A,B). The sculpturing does not cover lateralmost sides of the dorsum, legs and ventral body side ( Figure 21A,B,D). All legs with sensory organs: poorly marked and weakly protruding papillae on legs I and II (Figures 21A, 23A); thin and spine-like on legs III ( Figure 21A); large papillae on legs IV of the shape analogous to primary clavae.

744
P. Gąsiorek and R. M. Kristensen Adult males. Body elongated ( Figure 21C), but qualitatively like in females beside of the circular gonopore ( Figure 23C).

Cuticular sculpturing in Echiniscoidea
The utility of the cuticular ultrastructure as a systematic criterion has been demonstrated several times for both Heterotardigrada and Eutardigrada. In the latter, the most striking example is that of the Murrayidae, which share the plesiomorphic presence of pillars with Heterotardigrada (Kristensen 1982;Guidetti et al. 2000), whereas the remaining Macrobiotoidea are characterised by the apomorphic lack of pillars in cuticle. In contrast to the great majority of eutardigrades with a smooth cuticle, heterotardigrade cuticle typically contains pillars (Greven 1975;Kristensen & Neuhaus 1999) that may be particularly elongated, e.g. in the genera Raiarctus  and Rhomboarctus (Hansen et al. 2003). Dorsum is typically sculptured in representatives of Echiniscoidea: it consists of both endocuticular pillars and epicuticular granules/ tubercles in Carphaniidae, Oreellidae and Isoechiniscoidinae (Binda & Kristensen 1986;Dastych et al. 1998;Møbjerg et al. 2016). Echiniscoidinae lack pillars  and sometimes their cuticle is completely smooth Chang & Rho 1998), whereas sculpturing may be variously developed in Echiniscidae, frequently forming richly ornamented patterns (e.g. Kristensen 1987;Guil et al. 2013;Gąsiorek et al. 2019). The interspecific variability mirrored in dorsal sculpturing of Echiniscoididae is so pronounced (e.g. compare E. sigismundi ( Figure 24A,B) with Neoechiniscoides horningi (Miller & Kristensen, 1999) and an undescribed species from Pacific Ocean ( Figure 24C)) that, after four decades since the first echiniscoidid revision , we decided to elevate all subspecies of E. sigismundi to species level. This move is supported by significant disparities in claw numbers, cuticle morphotypes and geographic isolation of these taxa as well (only E. groenlandicus + E. sigismundi and E. porphyrae + E. verrucariae were found in the same localities). Some former  . The division of the family Echiniscoididae into two subfamilies based on several morphological criteria, including the development of cuticular pillars, is firmly supported in our COI-based phylogeny (Figure 1), and further corroborates the usefulness of pillars in tardigrade classification (Møbjerg et al. 2018). In the future, when the splitting of Echiniscoides into new genera advances, it will be intriguing to investigate whether main sculpturing morphotypes (e.g. punctuations vs large polygons) correspond with molecularly reconstructed clades and could be introduced into putative diagnoses of newly erected echiniscoidid taxa.

Taxonomic key to Echiniscoididae
Summary of ecomorphological traits pertaining solely to Echiniscoides is shown in Table IX. The key presented below deals with traits observable in sexually mature (adult) specimens of all Echiniscoididae, and substitutes the narrow key from Kristensen and Hallas (1980). Given the wide variability in claw formula in some species, we used that trait as a last resort in certain cases.